Type II diabetes status, adipokines, and dioxin-like compounds in the Anniston Community Health Survey II

Abstract

Background/Aim: A polychlorinated biphenyl (PCB) production facility operated in Anniston, AL for several decades. Serum PCB concentrations were found to be about 3 times higher than the general U.S. population, and the Anniston Community Health Survey (ACHS 2005-07) found significant associations between PCBs and type II diabetes (T2D) prevalence. We also noted an inverse relationship between insulin and dioxin-like compounds in a subset of participants where the novel liver damage biomarker – cytokeratin 18 (CK-18) indicated non-alcoholic toxicant-associated steatohepatitis (TASH). Using data on serum concentration of dioxin-like compounds collected in the 2014 follow-up study (ACHS II), we investigated the association between dioxin-like compounds, insulin, measured cytokines, and adipokines as related to T2D status. Methods: T2D status was defined as previous physician diagnosis, measured glucose >125 mg/dL, or being on any glycemic medication; 135 (39.9%) of 338 participants were classified as diabetic. Adipokines were measured on HADK2MAG-61K and HADK1MAG-61K bead arrays (EMD Millipore, Billerica, MA). The polychlorinated dibenzo-p-dioxins (PCDD), dibenzofurans (PCDF), and non-ortho PCBs were measured using high-resolution gas chromatography/high-resolution mass spectrometry and expressed as dioxin toxic equivalents (TEQs, pg/g lipid). Linear regression models adjusted for age, sex, ethnicity, BMI, family history of diabetes, and smoking status.Results/Conclusion: We found significant associations between total dioxin concentration and insulin, leptin, and adiponectin for the total 338 participants. For insulin and leptin, this was primarily driven by PCDD/PCDF TEQ in those with diabetes (β=-0.31, p-value 0.034; β=-0.010, p-value 0.003). For adiponectin, these associations were stronger in those without diabetes (β=0.27, p-value 0.017). Participants with TASH saw a similar inverse relationship between PCDD/PCDF and insulin, and those with both diabetes and TASH had the largest measure of effect (β=-1.08, p-value 0.01). These findings suggest that these chemicals may affect concentrations of compounds regulating glucose levels as well as lipid metabolism and energy balance.